Tuesday, October 17, 2017

An Oracle Analysis

I'm going to illustrate how I interpret my Oracle results, because I still see a lot of people asking "what do my Oracle results mean?" If you haven't already, you may want to read my intro guide to Gedmatch's Admixture and Oracle, but I'd like to elaborate on that a little bit.

Firstly, it's important to remember that the results can be very speculative and it's best not to take them very literally. People in neighboring regions simply share too much DNA to always be able to tell them apart with accuracy. That means the more narrowed down the areas are in the result, the more speculative it is. You could be German, for example, and get French results because they are neighboring countries who share a lot of DNA. It doesn't mean you're French, it just means this particular calculator put that French/German shared DNA into French instead of German.

Eurogenes K13 Oracle 4, using 4 populations approximation
Secondly, your results are going to be different for each calculator you use so don't just stick to one, explore all those which apply to your background (ie, don't go using Ethiohelix when you're 100% European). Certain calculators may give you more or less accuracy than others. In my personal experience, Eurogenes K13 Oracle 4 (right) isn't very accurate. It really wants me to be Jewish and I'm really not - I have no known Jewish ancestry and don't get any Jewish results from any of the big 4 companies, or in any of Gedmatch's Admixtures. It crops up in the odd Oracle result, but none so much as Eurogenes K13 Oracle 4 populations. I personally have found K15 and EUtest Oracle's to be more accurate, and since K15 is a more recent version of EUtest, that's what I'm going to use to demonstrate how to read Oracle results in some more depth than before.

I find the best thing to do is rather than look at your Oracle results and try to pick one combination that fits you best, or shows the closest distance, look at the results on the whole. Which populations are you seeing the most? Which ones the least? Although I like to look at 4 populations the most because I am primarily from 4 different regions in Europe, you can also look at the 1, 2, and 3 populations modes.

These are my Eurogenes EUtest V2 K15 Oracle 4 results:

1 Orcadian + South_Italian + West_German + West_German @ 4.425306
2 French + South_Italian + West_German + West_Norwegian @ 4.689746
3 South_Italian + Southwest_English + West_German + West_German @ 4.689806
4 East_Sicilian + Orcadian + West_German + West_German @ 4.747531
5 Italian_Jewish + Orcadian + West_German + West_German @ 4.835878
6 East_Sicilian + Southwest_English + West_German + West_German @ 4.850750
7 Italian_Abruzzo + West_German + West_German + West_German @ 4.853912
8 North_Dutch + South_Italian + West_German + West_German @ 4.863277
9 French + Orcadian + South_Italian + West_German @ 4.911067
10 South_Italian + Southeast_English + West_German + West_German @ 4.914701
11 Tuscan + West_German + West_German + West_German @ 4.922722
12 Central_Greek + Orcadian + West_German + West_German @ 4.922800
13 South_Italian + West_German + West_German + West_Norwegian @ 4.927629
14 Irish + South_Italian + West_German + West_German @ 4.941526
15 South_Italian + West_German + West_German + West_Scottish @ 4.958009
16 East_Sicilian + French + West_German + West_Norwegian @ 4.978409
17 East_Sicilian + French + Orcadian + West_German @ 4.982550
18 South_Italian + West_German + West_German + West_German @ 5.005996
19 South_Italian + Spanish_Galicia + West_Norwegian + West_Norwegian @ 5.010231
20 Central_Greek + Southwest_English + West_German + West_German @ 5.011045

So rather than saying the top results must be the most accurate because it's the closest distance, and determining it to be only somewhat accurate because it did identify my German, Italian, and Scottish ancestry, but not my English or Norwegian, let's look at the entire results as a whole.

Map showing my known ancestor's birth
places in Europe
What am I seeing the most? Probably West German. This is very accurate, I have a lot of West German ancestry on both sides of my tree, and I estimate it makes up about 25% of my tree. I also have a couple Swiss-German branches, which is still fairly consistent with West German. There's one ancestor who was from Bavaria, which is a region of Germany more to the east, but I have no idea what part of Bavaria - could have been the western most part for all I know. What I do know is that I rarely ever get admixture/ethnicity results in Eastern Europe and when I do, it's normally in such small portions, it's likely noise. So this is all very consistent with my tree.

I'm also seeing South Italian and some other Italian regions like East Sicilian and Abruzzo. This is incredibly accurate. I do indeed have Sicilian, Abruzzo, and other Southern Italian ancestry. My paternal grandmother was of entirely Italian descent so that makes up another 25% of my tree. My Sicilian branch is a bit more Northern Sicily than Eastern, but that's fairly negligible. There's one count of Tuscan and as far as I know I have no Tuscan ancestry, but that too is probably not very significant since it only shows up once.

There's a few West Norwegians thrown in there, which is also accurate, I have one great grandparent who was Norwegian, making up 12.5% of my tree. Several branches were indeed from Western Norway, although one branch did come from the more Eastern towns of Bamble and Skien.

Map showing my population results for
Eurogenes K15 Oracle 4, compare with above map
You may also notice a few Orcadian and West Scottish populations. This is somewhat accurate, I do have several Scottish or Scots-Irish branches dating back to colonial times, but where exactly in Scotland they were from isn't really known. Orcadian (people from the Orkney Islands) seems a little unlikely as my understanding is most Orcadian immigrants went to Canada through the Hudson Bay Company rather than the US. But if we consider Orcadian as a representation of my Scottish or even British heritage, it makes sense. The Orcadians were also influenced by the Vikings, so there's also a potential connection to my Norwegian side. My overall British branches make up about 34% of my tree, and in addition to Scottish, includes English, so it's not surprising to find a few instances of Southeast or Southwest English. As you can see from the map above, I do indeed have ancestry in Southwest and Southeast England, although I have more recent roots in Northern England, near Manchester so it's a shame it didn't pick this up. It's possible Oracle is underestimating my British ancestry, since there's only a few English populations included, but when you consider how genetically similar the British and Germans are, and knowing how many instances of West German are listed, it makes some sense.

I lastly have a smidgen of colonial Dutch and French Huguenot in my tree but I don't know how realistic it is to expect that to show up in admixture results, as it may have been from too long ago. They make up about 1-2% each of my tree. So when I see a few results for North Dutch and French, I'm taking it with a grain of salt. I'd like to think it could be from colonial ancestry but the way Oracle works by identifying the populations you match most closely, it doesn't seem likely I would closely match a population from so far back in my tree. It seems more likely that it's just being picked up from neighboring regions where I have ancestry.

Likewise, I wouldn't put much thought into the remaining few instances of Italian Jewish, Central Greek, Irish, and Spanish Galicia. Irish is probably just representing my British background, as those two groups are closely related, and likewise, Italian Jewish, Central Greek, and Spanish Galicia may be related to my Italian heritage since they're all from that Mediterranean area. In any case, since there's only one or two counts of them, it's easy to ignore them.

So overall, despite the fact that it doesn't always identify my British/English ancestry as much as it maybe should, it's actually remarkably accurate when you look at it on the whole. Compare the two maps above, one showing the origins of my ancestors and the other showing my Oracle results, and they really aren't far off each other (keeping mind some of the locations for the Oracle map cover a larger area than what the pinpoints represent). Mapping it out is another good, fun way to analyze your admixture or Oracle results, if you'd like to try it, just go to My Google Maps.

I don't normally worry too much about the distance unless it starts getting really high. For example, K13's Oracle has closer distances than K15, but the populations in K15 are far more accurate for me than K13. I am NOT saying K15 is the best option for everyone. When I look at my dad's K15 Oracle results, they are mostly inaccurate, constantly insisting he is Lebanese Druze, which seems very off base. I can't even promise that there will be an Oracle calculator that is as accurate for you as Eurogenes K15 is for me, since I haven't really found one for my dad that is this accurate (he does get a lot of Abruzzo results in various calculators, which is accurate, but there's also a lot of populations that are kind of out there for him).

Also keep in mind some of the calculators contain a lot of ancient (prehistoric) populations. If you see some weird names like "Battle Axe" or "Bell Beaker", these are probably ancient populations (Battle Axe is Neolithic, Bell Beaker is western Europe in late Neolithic-early Bronze Age).

I hope this gives some more detailed insight in how you might interpret your own Oracle results. If you are adopted and don't know your ancestral background, it's difficult to know which calculators will be more accurate than others. You should definitely still take all this with a grain of salt, but it is fun to examine and compare with what we do know.

Wednesday, October 4, 2017

The Forgotten Witch Trials of Connecticut, 1647-1697

Dramatized depiction of a witch trial
I recently discovered that I had an ancestor involved in the witch trials of Connecticut and of course immediately went to look up more information on this subject. I was very surprised to find that while there are a lot of articles about it around the internet, none came from the popular Wikipedia. (Edit: this has finally now changed - see here). There are only a few books which detail the Connecticut events and fewer still which are dedicated entirely to them. Even history buffs often admit to not knowing about the Connecticut witch trials, in spite of the fact that the very first ones in the colonies occurred in Connecticut. It's safe to say they are greatly eclipsed by the Salem witch trials, which perhaps receive more attention because they occurred over a much shorter time period. Salem was very much a frenzied hysteria with the executions of 20 people within just over a year (February 1692 to May 1693), whereas the trials in Connecticut resulted in 35 cases and just 11 executions over the course of 50 years (1647-1697). Salem certainly deserves the attention it gets, but it should not be at the cost of forgetting other important witch hunts too.

Oddly, the ancestor in question, Christopher Comstock, has his own Wikipedia page, despite the greater trials in which he was involved not having one. Christopher was involved in the witch trials twice, firstly in 1653-1654 when he gave an affidavit about having visited Goodwife Knapp while she was in prison for witchcraft. Knapp was later executed. Secondly, he served on the grand jury investigating witchcraft in Connecticut in September 1692.

One of the reasons these trials kept cropping up was because every time someone was accused of witchcraft, they were pressured to "confess" and name others they knew of who were also witches. According to the author of "The witchcraft delusion in colonial Connecticut," from the moment Knapp was sentenced she "was made the object of rudest treatment, espionage, and of inhuman attempts to wring from her lips a confession of her own guilt or an accusation against some other person as a witch." Just as we might question a terrorist to confess who they are working with, this logic was applied to "witches" too in the 17th century. This is where my ancestor Christopher Comstock comes in. In 1653, Goodwife Knapp, whose first name is lost to history, was in prison in Fairfield for witchcraft. Comstock, along with Thomas Sheruington and Goodwife Baldwin, visited her in her cell where Baldwin questioned her about her fellow "witches". It sounds as though Comstock and Sheruington were merely there as witnesses. Knapp admitted that she knew some, or at least one person who had "received Indian gods that were very bright." Knapp was claiming her innocence so Baldwin asked her how she could know this if she weren't a witch herself, to which Knapp responded that the guilty party had told her so. It appears that Knapp did not reveal the name of the person who told her this though. During another questioning by Mistress Pell, Knapp insisted, "I have sins enough already, and I will not add this [accusing another] to my condemnation."

The court didn't believe her plea of not-guilty, because Knapp was convicted and executed by hanging. She went to the grave pleading her innocence. My ancestor's role in this was minor, he was merely witness to an interview with Knapp as prisoner. His affidavit was not even used at her trial, since it was actually written after the fact, to be used in another case the following year. Unfortunately, there are few details about Knapp's trial, we do not even know the specifics of what she was accused of, who accused her, what the testimonies against her included, etc. Most of what we know about Knapp comes from an investigation after her 1653 execution in which testimonies were given about Knapp's supposed accusations of another, Mary Staples, which is when Comstock wrote his affidavit.

After Knapp's execution, her body was desecrated when several individuals stripped it and searched it for marks of a witch. Mary Staples proclaimed there were no marks on Knapp's body that couldn't also be found on herself, an attempt to claim there were no witch's marks on Knapp's body. Later, Robert Ludlow claimed that just before her execution, Knapp had requested to speak to him privately, during which she told him that Mary Staples was a witch. This seems unlikely given the fact that she wouldn't name anyone under extreme pressure and duress in her cell. Why would she suddenly, on her own accord, decide to accuse Mary Staples, and furthermore, why would she do so privately, with no witnesses, if she wanted it known? It's believed Ludlow took Mary's comments not to mean Knapp had no witch's marks, but that both Knapp and Mary had them and that made Mary a witch too. But the conflict between Ludlow and Staples had been going on since at least 1651 when Ludlow won a law suit against Mary for slander, so Ludlow was likely looking for anyway he could to accuse her of anything else. Mary's husband, Thomas Staples, caught wind of Ludlow's tale, and in attempts to forestall the accusations against his wife, brought suit against Ludlow in 1654 for defamation of character, and there began the investigation in 1654, including Comstock's affidavit. There was also a witness account given by another of my ancestors, Rose Sherwood, then the wife of Thomas Barlow. Rose testified that after Knapp's execution, she was among those women who searched Knapp's body for marks. She claims at first they found nothing unusual, but then upon another look, they did.

Despite several testimonies against Mary Staples, in the end, the court saw reason and ruled in favor of her husband, awarding Ludlow with damages for defamation of character. It did not prevent a later trial against Mary for witchcraft though, in 1692, but Ludlow had left Connecticut by then and Mary was fortunately acquitted.

It is relieving to see that Comstock's affidavit did not contribute to any conviction or execution. He was merely an observer, witness of something Knapp had said, which was later used by others as an attempt to accuse someone else, but it failed. It's hard to say what he thought or felt about it. Comstock is believed to have been born about 1635, which would have made him only 18 at the time he witnessed the questioning of Knapp in 1653. If that's the case, he was quite young and his experiences in these trials must have helped shaped his development into an adult.

What else is known of Knapp is very little. In John Taylor's "The witchcraft delusion", all it says of Knapp herself is that she was "presumably a woman of good repute, and not a common scold, an outcast, or a harridan" and quotes other sources saying "she impresses one as the best woman" and that she was a "just and high minded old lady."

John Winthrop Jr.
Fast forward to 1692. Salem is in its height of witch trial hysteria and Connecticut isn't far behind, with the trials of six women in Fairfield, all accused by the same servant girl, Katherine Branch. Fortunately, unlike in Salem, none were executed. After Hartford saw the trials of nine people and the executions of four of them in 1662, the Connecticut governor John Winthrop Jr made it necessary for two witnesses for each alleged act of witchcraft to be required for a conviction, rather than only one. This made convictions much more difficult and resulted in no further executions of witches in Connecticut after 1662. Winthrop appears to have been the saving grace of Connecticut, and something of an antithesis to Salem's Cotton Mather, often personally overturning or reversing convictions. He may have been the main reason Connecticut had fewer executions of witches overall than Salem, and none at all during Salem's mass of them in 1692. However, that's not to say the Fairfield trials in 1692 didn't results in any convictions at all. Of the six women accused, three were acquitted, two never went to trial (jury found no bill, meaning there wasn't enough evidence to go to trial), and one, Mercy Disborough, was convicted. She was never executed though, as she was later pardoned. Christopher Comstock was on the jury that convicted Mercy, but also acquitted and found no bill for the other five women. So my ancestor was (partially) responsible for the conviction of Mercy Disborough on the charge of witchcraft, but fortunately not for her death.

An engraving of one floating on water
during ordeal by water (ie, guilty)
Although Katherine Branch made the initial accusation, there were numerous testimonies against Mercy, so it seems she ruffled more than enough feathers, though nothing that should warrant her execution. Most of the accusations were ridiculous to think they could be related to Mercy, including one unnamed young woman prone to seizures who accused Mercy of being responsible for them. Mercy was subjected to being searched naked for marks of the devil, and even to the water test, or ordeal by water. This is the notorious test where one's hands and feet are bound together before being thrown into the water and if they sink, they are considered innocent, and if they float, they are considered guilty. The basis of this was the ridiculous theory that witches floated because they had renounced baptism and therefore were being rejected by the water. Another idea was that witches were supernaturally light weight. In any case, naturally, they were pulled out of the water before they drown, by a rope which was tied to them. The idea that this sort of test meant the individual on trial would die whether found innocent or guilty (drown if innocent, executed if guilty) is a modern misconception. Mercy, along with another accused (Elizabeth Clawson), were tied up and thrown into the water on September 15, 1692, where two witnesses (Abram Adams and Jonathan Squire) claimed they floated like corks, and even when pressed down into the water, they bounced back up. However, this test obviously wasn't the deciding factor in the trials, since although Mercy was convicted, Elizabeth was not, despite both of them floating. That suggests enough people at the time were skeptical of the authenticity of such a test that its results weren't taken into great consideration.

Apart from Mercy Disborough and Elizabeth Clawson, the others who were on trial in Fairfield in 1692, accused by Katherine Branch (a servant of Daniel Wescot/Westcott), included: Mary Harvey, Hannah Harvey, Goody Miller, and Mary Staples, the same Mary Staples whose husband sued Robert Ludlow for defamation of her character and won. Most of the other Connecticut cases took place in other towns, including Windsor, Hartford, Wethersfield, New Haven, East Hampton, Saybrook, Stratford, and Wallingford, though some of them were tried in Hartford instead.

Although the Connecticut cases were spread out over time and saw fewer executions than Salem, they still played an important role in the history of witch trials and should not be forgotten.


Also check out:
  • Connecticut Witch Trials: The First Panic in the New World by Cynthia Wolfe Boynton  
  • Before Salem: Witch Hunting in the Connecticut River Valley, 1647–1663 by Richard S. III Ross
  • Escaping Salem: The Other Witch Hunt of 1692 (New Narratives in American History) by Richard Godbeer 
  • Witchcraft Trials of Connecticut by Richard G. Tomlinson